Small Value: 15 years of underperforming

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Dead Man Walking
Posts: 906
Joined: Wed Nov 07, 2007 6:51 pm

Re: Small Value: 15 years of underperforming

Post by Dead Man Walking » Sat Nov 24, 2018 9:19 pm

These scv threads always compare the performance of various indices or funds tracking them. DFA funds are usually included in the discussion. A few Bogleheads view DFA funds as active funds. Many believe that backtesting should be limited to retail funds that are actually performing in the real world. I don't recall a thread that compared the performance of all scv funds with the performance of all scg funds. By "all" - I mean active and passive funds. Of course, that would be an arduous task that would present survivorship bias problems. However, the fund universe could be screened for funds in existence during various time periods.

DMW

fennewaldaj
Posts: 823
Joined: Sun Oct 22, 2017 11:30 pm

Re: Small Value: 15 years of underperforming

Post by fennewaldaj » Sat Nov 24, 2018 10:52 pm

Robert T wrote:
Sat Nov 24, 2018 4:32 pm
.


Most, if not all, small growth funds (etfs) are not the opposite of value, while the FF small value and growth series are exact opposites e.g.

FF Value = high price to book,
FF Growth = low price to book (no other metrics included).

Compare this for example to the S&P value and growth sorts

S&P Value = High book-to-price, earnings-to-price, sales-to-price.
S&P Growth sorts do not use the same metrics at the opposite end of the spectrum e.g. low book-to-price, earnings-to-price, sales-to-price.
S&P Growth = Three-year changes in earnings per share over price per share, three-year sales per share growth rate, and momentum.

This leads to much confusion in the discussion with, it seems an assumption, that all funds/etfs with the name “growth” are the exact opposite to “value’ (same sort metrics but at opposite ends of the spectrum), similar to the FF growth and value series. However, the ‘growth’ funds often have completely different sort metrics (as in the S&P series above) and are not opposite to “value”. This is why some “growth” funds have a positive value load as in the case of the S&P600 Small Growth.

What matters over the long haul is factor exposure and expense. The rest is largely noise.

Obviously, no guarantees over any time period (including for the equity premium).

Robert
.
Would it be fair to say small growth as it appears in mutual funds is much better investment than small growth as defined by FF? It seems like none of the funds are really investing in the black hole stocks that make FF SG look so bad.

heyyou
Posts: 3597
Joined: Tue Feb 20, 2007 4:58 pm

Re: Small Value: 15 years of underperforming

Post by heyyou » Sun Nov 25, 2018 12:43 am

Please consider how most allocations have done quite well over time. This debate has been about what has done better or best, when all have done well enough for us Main Street investors. The performance variations could be absorbed by our choosing of when or to where we retire, not about if we ever get to retire. Just be grateful that we each do have adequate returns, instead of constantly (and greedily) seeking an elusive optimal.

We could each outperform those small portfolio performance differences by just saving a little more, spending slightly less as retirees, or bumping our equity exposure by several percentage points.

Park
Posts: 691
Joined: Sat Nov 06, 2010 4:56 pm

Re: Small Value: 15 years of underperforming

Post by Park » Sun Nov 25, 2018 12:45 am

I do notice the shortfall between value indices and value funds. That does not make me optimistic about the new multifactor funds.

User avatar
Robert T
Posts: 2610
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small Value: 15 years of underperforming

Post by Robert T » Sun Nov 25, 2018 5:46 am

fennewaldaj wrote:
Sat Nov 24, 2018 10:52 pm
Would it be fair to say small growth as it appears in mutual funds is much better investment than small growth as defined by FF? It seems like none of the funds are really investing in the black hole stocks that make FF SG look so bad.
Yes - index funds that call themselves "small growth" (as well as their historical series) show better annualized returns than the FF Small Growth series. I don't think it is attractive for a mutual fund provider to base a fund exclusively on the FF Small Growth series. What would the sales pitch be?

The 'small growth' funds seems to be more a blend of the FF Small Growth and FF Small Neutral series, and likely explains some of the recent (past 15 years) difference in the returns between FF Small Growth and "Small Growth" index funds. Since 2003 the annually rebalanced FF Small neutral series had a higher return than both the annually rebalanced FF Small Growth and FF Small Value Series. As a result a "growth fund" heavily tilted towards FF Small Neutral (e.g. 20% FF Small Growth:80% FF Small Neutral) since 2003 had higher return than the FF Small Value series as observed with "small growth" index funds vs. "small value" index funds.

That all being said, "small value" has still done as well as, or outperformed, "small growth" in the longer historical index series e.g.

Annualized return since index inception:

7/2001 - 10/2018
CRSP Small Value = 10.3
CRSP Small Growth = 8.4

1/1979 - 10/2018
Russell 2000 Value = 12.8
Russell 2000 Growth = 9.8

7/1995 - 10/2018
S&P SmallCap 600 Value = 11.0
S&P SmallCap 600 Growth = 10.9*

*The way the S&P SmallCap 600 "Growth" series kept up with the "Value" series was to itself have a value tilt (+ momentum). The value load on the "Growth" series was +0.18 for this period.

Obviously no guarantees.

Robert
.

betablocker
Posts: 504
Joined: Mon Jan 11, 2016 1:26 pm

Re: Small Value: 15 years of underperforming

Post by betablocker » Sun Nov 25, 2018 9:35 am

JBTX wrote:
Thu Nov 22, 2018 11:20 pm
stlutz wrote:
Thu Nov 22, 2018 8:07 pm
Okay, so here is a comparison of Vanguard Small Growth and Value with DFA Small Value thrown in from 6/98 through 9/18:

https://www.portfoliovisualizer.com/bac ... ion3_3=100

Results:

VG Small Value: 9.06%
VG Small Growth: 9.45%

DFA Small Value: 9.87%

I pulled down the monthly returns from the Ken French data library (http://mba.tuck.dartmouth.edu/pages/fac ... brary.html), I get the following annualized returns for the same period:

Small Value: 12.28%
Small Growth: 7.58%

So, this is the problem with the whole smallcap value thing: Theoretically over this period the value premium in smallcaps was 4.70% per year. That's quite large. The actual premium when using index funds was -.39%.

That's the problem with the big premiums people are looking for by tilting to value in the small cap face. The theoretical numbers simply don't correlate to what real-world mutual funds are investing on (or can invest in). In fact, if you run a factor regression on Vanguard Small Growth, it has a modestly positive loading on value!

A fair number of people believe that the value premium is much larger in small caps than it is in large caps. This appears not to be the case--this result is based on more of a quirk in how the databases were used as opposed to representing real world return differences. In the real world, there is no small cap growth "black hole".
I'd like to know what caused the difference in the French data and index returns of anybody knows.
I believe the FF factor returns are long/short. The funds at dfa and vanguard are long only. Additionally you have some issues with how much small and value a fund can tap. With massive aum vanguard has to buy bigger stocks with less value load.

Topic Author
Elysium
Posts: 1985
Joined: Mon Apr 02, 2007 6:22 pm

Re: Small Value: 15 years of underperforming

Post by Elysium » Sun Nov 25, 2018 10:04 am

Robert T wrote:
Sun Nov 25, 2018 5:46 am
The 'small growth' funds seems to be more a blend of the FF Small Growth and FF Small Neutral series, and likely explains some of the recent (past 15 years) difference in the returns between FF Small Growth and "Small Growth" index funds. Since 2003 the annually rebalanced FF Small neutral series had a higher return than both the annually rebalanced FF Small Growth and FF Small Value Series. As a result a "growth fund" heavily tilted towards FF Small Neutral (e.g. 20% FF Small Growth:80% FF Small Neutral) since 2003 had higher return than the FF Small Value series as observed with "small growth" index funds vs. "small value" index funds.
Isn't this the problem then with relying on FF research and data, which is more and more looking like an academic excercise while real implementations are nowhere near whether the funds are called value, growth, or blend. Based on this, one could conclude that funds that are labeled value, growth, and blend are likely to have similar returns over long periods, while short term variations could show growth outperforming value and vice versa.

That is what Jack Bogle concluded in his early book, and cited fund data to show value and growth tend to have similar returns over time. Dr. Bill Bernstein, who incidentally posted on this thread asking whether SV orange has anymore juice left in it, said once that (paraphrasing) when you disagree with Jack Bogle you end up with egg in your face.

Random Walker
Posts: 4233
Joined: Fri Feb 23, 2007 8:21 pm

Re: Small Value: 15 years of underperforming

Post by Random Walker » Sun Nov 25, 2018 11:41 am

Elysium wrote:
Sun Nov 25, 2018 10:04 am
Robert T wrote:
Sun Nov 25, 2018 5:46 am
The 'small growth' funds seems to be more a blend of the FF Small Growth and FF Small Neutral series, and likely explains some of the recent (past 15 years) difference in the returns between FF Small Growth and "Small Growth" index funds. Since 2003 the annually rebalanced FF Small neutral series had a higher return than both the annually rebalanced FF Small Growth and FF Small Value Series. As a result a "growth fund" heavily tilted towards FF Small Neutral (e.g. 20% FF Small Growth:80% FF Small Neutral) since 2003 had higher return than the FF Small Value series as observed with "small growth" index funds vs. "small value" index funds.
Isn't this the problem then with relying on FF research and data, which is more and more looking like an academic excercise while real implementations are nowhere near whether the funds are called value, growth, or blend. Based on this, one could conclude that funds that are labeled value, growth, and blend are likely to have similar returns over long periods, while short term variations could show growth outperforming value and vice versa.

That is what Jack Bogle concluded in his early book, and cited fund data to show value and growth tend to have similar returns over time. Dr. Bill Bernstein, who incidentally posted on this thread asking whether SV orange has anymore juice left in it, said once that (paraphrasing) when you disagree with Jack Bogle you end up with egg in your face.
I don’t think there is a problem with relying on FF data. I think the problem is with relying on the advertised name of some funds and finding passive funds that stick to a defined factor mandate. The FF data crystallize what factors have been shown to be significant, and it’s up to the investor to apply the information as he will. That requires appreciating the true factor exposure of the funds chosen for actual investment.

Dave

Topic Author
Elysium
Posts: 1985
Joined: Mon Apr 02, 2007 6:22 pm

Re: Small Value: 15 years of underperforming

Post by Elysium » Sun Nov 25, 2018 12:02 pm

Random Walker wrote:
Sun Nov 25, 2018 11:41 am
I don’t think there is a problem with relying on FF data. I think the problem is with relying on the advertised name of some funds and finding passive funds that stick to a defined factor mandate. The FF data crystallize what factors have been shown to be significant, and it’s up to the investor to apply the information as he will. That requires appreciating the true factor exposure of the funds chosen for actual investment.
I think we are going in circles here. It makes no sense to blame SCG funds of not picking the so called black hole stocks in FF SG universe, and credit to them for picking the neutral, momentum, profitability, or any number of other factors that are positive. Which then makes this whole point (of SCV) moot. We can conclude that any number or classifications of funds could be implementing a number of these factors and any one of them could achieve similar results over long periods. Leaving random variations in between as when a particular factor is in favor or not.

In other words, if we buy the whole basket, you get it all.
Last edited by Elysium on Sun Nov 25, 2018 12:04 pm, edited 2 times in total.

garlandwhizzer
Posts: 2572
Joined: Fri Aug 06, 2010 3:42 pm

Re: Small Value: 15 years of underperforming

Post by garlandwhizzer » Sun Nov 25, 2018 12:03 pm

stlutz wrote:

Okay, so here is a comparison of Vanguard Small Growth and Value with DFA Small Value thrown in from 6/98 through 9/18:

https://www.portfoliovisualizer.com/bac ... ion3_3=100

Results:

VG Small Value: 9.06%
VG Small Growth: 9.45%

DFA Small Value: 9.87%

I pulled down the monthly returns from the Ken French data library (http://mba.tuck.dartmouth.edu/pages/fac ... brary.html), I get the following annualized returns for the same period:

Small Value: 12.28%
Small Growth: 7.58%

So, this is the problem with the whole smallcap value thing: Theoretically over this period the value premium in smallcaps was 4.70% per year. That's quite large. The actual premium when using index funds was -.39%.

That's the problem with the big premiums people are looking for by tilting to value in the small cap face. The theoretical numbers simply don't correlate to what real-world mutual funds are investing on (or can invest in)
To me this is the essence of the problem with factor investing. Backtesting uses long/short portfolios that are cost free. No such thing exists in real funds. Shorting small cap stocks is a total loser's game, costs overcome any potential benefit. So by removing the short position, the so called premium is cut in half automatically. Trading stocks in the more illiquid SC space is inefficient due to low trading volumes. When you buy a significant portion in a SCv stock the price of the stock goes up as you buy more, and the opposite happens as you try to sell it. Cost free indexes do not reflect this inefficiency, trading is assumed to be free (wrong) and not to affect trading price (also wrong). So the premium (which has already been cut in half by no shorting) is further reduced by trading frictions and costs. Then we add to the cost structure the increased fund fees/charges that must be paid to those who create and market these products which subtracts more from the supposed alpha. Finally we must take into effect that many SCV quants are looking at the same data and pickling the same SCV stocks to buy (distorting their prices in the wrong direction and reducing the very thing they seek, value). So after it's all said and done, how is the investor in the average SCV value fund going to do? That is the important question, not whether there is academic research for cost-free long/short portfolios, or convincing narratives for the existence of the SCV premium.

Academics get promotions and often high paying jobs from private factor based investing companies for cranking out convincing factor research. Those who market and sell such products also get handsomely rewarded. There are built-in generous incentives to producing what appears to be a better mouse trap, hence the proliferation of factors and approaches to harvest them. Whether or not the better mouse trap actually works for investors is much less clear than the academic research might suggest.

One final note. All premiums except one are based on cost-free long/short portfolios which, as I have pointed out, is an invalid measuring tool. The exception? Beta. No short is used to double the the expected return of long only beta, except possibly the risk free rate, T Bills which in real terms inflation-adjusted is rarely more than zero. On the other hand cost-free short equity portfolios is often theoretically quite positive in real terms. So when you're comparing expected returns of the the market versus factors it seems reasonable in order to make a valid comparison to almost double the expected return of the market factor in order to level the playing field. Very few real factor funds use long/short portfolios which says something about how shorting works in the real world after costs. Those that do short tend to have inconsistent results. Shorting works if you know which stocks are going to decline in value by a given date in the future and by how much it will decline. The problem is: no one knows that up front on a consistent basis as QSPIX's recent results demonstrate.

Garland Whizzer

User avatar
Robert T
Posts: 2610
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small Value: 15 years of underperforming

Post by Robert T » Sun Nov 25, 2018 12:13 pm

Elysium wrote:
Sun Nov 25, 2018 10:04 am
Isn't this the problem then with relying on FF research and data, which is more and more looking like an academic excercise while real implementations are nowhere near whether the funds are called value, growth, or blend. Based on this, one could conclude that funds that are labeled value, growth, and blend are likely to have similar returns over long periods, while short term variations could show growth outperforming value and vice versa.
I think the FF research is even more important (part of the solution, not the problem) in understanding fund characteristic when the word “growth” is used for the name of funds that actually have value characteristics (e.g. S&P600 Small Growth). I would also not conclude that funds labeled value and growth are likely to have similar returns over long periods e.g. Since inception 8/2000 to 10/2018 annualized return of iShares Russell 2000 Value = 9.2% vs. iShares Russell 2000 Growth = 5.8%, and as in the earlier post since 1979, Russell 2000 Value Index = 12.8%, Russell 2000 Growth Index = 9.8%. “What matters over the long-haul is factor exposure and expense.” FF research helps us understand what factor exposure has been for various funds/indexes. I do think due diligence in this respect is important to target the factor exposure an investor wants, rather than random selection of something called value, or something called growth and hoping for the best.

Some investors may not believe in the value, momentum, and profitability premiums and want to just hold the market - and thats perfectly okay. Others may believe in these premiums and want some targeted exposure to them. FF research can help to do that in a more structured way than random selection based on fund names.

Obviously no guarantees.

Robert
.

User avatar
nedsaid
Posts: 12983
Joined: Fri Nov 23, 2012 12:33 pm

Re: Small Value: 15 years of underperforming

Post by nedsaid » Sun Nov 25, 2018 12:22 pm

Garland, the better indexes do a good job filtering out the "lottery stocks" the academics rail about. This is why when you use actual indexes to compare factor strategies against indexes that you often don't see the premiums. From all I have been reading, it seems that Quality/Profitability is where we should be looking. Quality/Profitability helps the Size premium as it does the Value premium. If I had to guess, it would probably help the Momentum premium as well. If you set your screening for Small, Value, and Quality, it would seem to me that many "value traps" would be screened out as well.

It might be that it is all about screening out the "anti-factors" more than screening for the factors themselves. Getting those "lottery stocks" and "value traps" out will boost the performance of about anything. I argue that the better indexes work well because the "lottery stocks" are just not there and that many of the "value traps" are not there either.

I think what is happening is that a single factor doesn't work so well anymore. You have to combine at least two now maybe three, or so it seems. The markets are probably getting more efficient even though more and more money is passively invested in indexes. Some smart aleck suggested that academics publishing studies has an adverse effect on the factors, if they would just stop publishing perhaps the factors would reappear.
A fool and his money are good for business.

Random Walker
Posts: 4233
Joined: Fri Feb 23, 2007 8:21 pm

Re: Small Value: 15 years of underperforming

Post by Random Walker » Sun Nov 25, 2018 12:58 pm

Elysium wrote:
Sun Nov 25, 2018 12:02 pm
Random Walker wrote:
Sun Nov 25, 2018 11:41 am
I don’t think there is a problem with relying on FF data. I think the problem is with relying on the advertised name of some funds and finding passive funds that stick to a defined factor mandate. The FF data crystallize what factors have been shown to be significant, and it’s up to the investor to apply the information as he will. That requires appreciating the true factor exposure of the funds chosen for actual investment.
I think we are going in circles here. It makes no sense to blame SCG funds of not picking the so called black hole stocks in FF SG universe, and credit to them for picking the neutral, momentum, profitability, or any number of other factors that are positive. Which then makes this whole point (of SCV) moot. We can conclude that any number or classifications of funds could be implementing a number of these factors and any one of them could achieve similar results over long periods. Leaving random variations in between as when a particular factor is in favor or not.

In other words, if we buy the whole basket, you get it all.
Another conclusion is simply it’s up to the investor to know what he is actually buying. Whether an investor wants deep value, value, neutral, growth, or total market is his choice. But in any of those cases, he should buy a passive fund that strictly adhered to its advertised factor exposure.

Dave

User avatar
hdas
Posts: 1337
Joined: Thu Jun 11, 2015 8:24 am

[Deleted]

Post by hdas » Sun Nov 25, 2018 1:21 pm

[Deleted]
Last edited by hdas on Wed Jan 29, 2020 12:10 pm, edited 1 time in total.
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

BigJohn
Posts: 1807
Joined: Wed Apr 02, 2014 11:27 pm

Re: Small Value: 15 years of underperforming

Post by BigJohn » Sun Nov 25, 2018 2:13 pm

randomguy wrote:
Sat Nov 24, 2018 5:42 pm
So the question isn't about SV performance against growth. It is about why the Small value mutual funds have not been able to match the fama-French portfolios? I am pretty sure if you compared the intersection of the stocks each held, it would be pretty clear with the average small value mutual fund being larger and less valuely.
Let's assume that the FF portfolio performance is real and persistent. The question for any potential investor then becomes how to move from the academic model to real world implementation. If the funds available (or at least under discussion in this thread) aren't small or valuely enough, then it's not surprising that they don't perform per the FF expectations. However, the question remains, which funds will match the expectations. If none do, then I would judge the FF model as interesting academic information which may not be practical for real world implementation.

I think a potential reason for this isn't hard to conceptualize. As your screen become smaller and more valuely, the total market cap available to invest in becomes smaller and smaller. Each company wants to offer a SV fund and as they attract more $$, they are forced to "expand" the screen or close the fund to new investors. As far as I can tell, none of the big players have closed their SV fund to new investors. This is analogous to the "too big to succeed" stories of some famous active funds (eg Magellan). The investment lesson I've learned in my reading is not to assume that these high performing ideas scale well. Just because something works at X size does not mean it will still work at 10X and it's more likely to be just the opposite. So even if you find just the right level of small and value in a fund today, how sure are you that this will still be the case in 5 or 10 years? This is why I decided not to tilt years ago, there is no worry about scaling or too big to succeed with a total stock index.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Small Value: 15 years of underperforming

Post by vineviz » Sun Nov 25, 2018 2:37 pm

BigJohn wrote:
Sun Nov 25, 2018 2:13 pm
randomguy wrote:
Sat Nov 24, 2018 5:42 pm
So the question isn't about SV performance against growth. It is about why the Small value mutual funds have not been able to match the fama-French portfolios? I am pretty sure if you compared the intersection of the stocks each held, it would be pretty clear with the average small value mutual fund being larger and less valuely.
Let's assume that the FF portfolio performance is real and persistent. The question for any potential investor then becomes how to move from the academic model to real world implementation. If the funds available (or at least under discussion in this thread) aren't small or valuely enough, then it's not surprising that they don't perform per the FF expectations.
I'm not sure the conclusion you've reached (that existing funds "aren't small or valuely enough") is necessarily the right one. All of the inexpensive (ER < 0.5%) small cap value funds with 10 year track records have statistically significant positive factor loadings on both size and value.

The diligence that investors should exercise comes in making sure these funds aren't ALSO exposed to other factors that the investor doesn't desire, and in making sure that when comparing the small cap value funds to other funds or to a benchmark that appropriate comparisons are made.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
privatefarmer
Posts: 609
Joined: Mon Sep 08, 2014 2:45 pm

Re: Small Value: 15 years of underperforming

Post by privatefarmer » Sun Nov 25, 2018 2:37 pm

Socrates28 wrote:
Thu Nov 22, 2018 7:12 pm
Vanguard small cap value index is much more midcap than small cap value
True but if you expect a “risk premium” then it should be positive, albeit to a lesser degree, amongst mid cap stocks as well as small. And I would argue vanguard is more half mid half small, making it smaller than straight mid cap. But you would expect to see outperformance if smaller stocks did better than large.

Look at the 30 year history of vanguards small fund vs TSM. It’s surprising. My conclusion is that one shouldn’t be so naive to think that it’d be so simple and easy to beat the market. EVERYONE now has access to small cap in their brokerage accounts, 401ks etc it’s so popular and widely available one would be foolish to think that the market would let it consistently outperform.

User avatar
nisiprius
Advisory Board
Posts: 39799
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Small Value: 15 years of underperforming

Post by nisiprius » Sun Nov 25, 2018 2:41 pm

betablocker wrote:
Sun Nov 25, 2018 9:35 am
...I believe the FF factor returns are long/short. The funds at dfa and vanguard are long only...
Let's be very clear on this point, because now that long-short funds are getting attention it gets confusing.

1) The Fama-French portfolios, like the ones I cited and St. Lutz cited before me, are long-only.
2) The Fama-French factors are long-short.

Up until a few years ago, most factor advocates were claiming great superiority portfolios with factor tilts obtained from long-only funds providing factor exposure through selection and concentration only.

Even today, even funds with names like "pure value" do not attempt to provide access to the pure value factor itself (i.e. using a long-short portfolio). I think that's a little curious. Is it because nobody wants one, or isn't it feasible?

Along the same lines, if nobody has a Fama-French small-cap value factor fund with a long-short portfolio, why doesn't anyone at least offer a "Fama French small-cap value index fund" that tracks the Fama-French small-cap value portfolio? Again: nobody wants one, or not feasible? And, if not, why not?
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
hdas
Posts: 1337
Joined: Thu Jun 11, 2015 8:24 am

[Deleted]

Post by hdas » Sun Nov 25, 2018 2:42 pm

[Deleted]
Last edited by hdas on Wed Jan 29, 2020 12:10 pm, edited 1 time in total.
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

User avatar
privatefarmer
Posts: 609
Joined: Mon Sep 08, 2014 2:45 pm

Re: Small Value: 15 years of underperforming

Post by privatefarmer » Sun Nov 25, 2018 2:45 pm

garlandwhizzer wrote:
Sun Nov 25, 2018 12:03 pm
stlutz wrote:

Okay, so here is a comparison of Vanguard Small Growth and Value with DFA Small Value thrown in from 6/98 through 9/18:

https://www.portfoliovisualizer.com/bac ... ion3_3=100

Results:

VG Small Value: 9.06%
VG Small Growth: 9.45%

DFA Small Value: 9.87%

I pulled down the monthly returns from the Ken French data library (http://mba.tuck.dartmouth.edu/pages/fac ... brary.html), I get the following annualized returns for the same period:

Small Value: 12.28%
Small Growth: 7.58%

So, this is the problem with the whole smallcap value thing: Theoretically over this period the value premium in smallcaps was 4.70% per year. That's quite large. The actual premium when using index funds was -.39%.

That's the problem with the big premiums people are looking for by tilting to value in the small cap face. The theoretical numbers simply don't correlate to what real-world mutual funds are investing on (or can invest in)
To me this is the essence of the problem with factor investing. Backtesting uses long/short portfolios that are cost free. No such thing exists in real funds. Shorting small cap stocks is a total loser's game, costs overcome any potential benefit. So by removing the short position, the so called premium is cut in half automatically. Trading stocks in the more illiquid SC space is inefficient due to low trading volumes. When you buy a significant portion in a SCv stock the price of the stock goes up as you buy more, and the opposite happens as you try to sell it. Cost free indexes do not reflect this inefficiency, trading is assumed to be free (wrong) and not to affect trading price (also wrong). So the premium (which has already been cut in half by no shorting) is further reduced by trading frictions and costs. Then we add to the cost structure the increased fund fees/charges that must be paid to those who create and market these products which subtracts more from the supposed alpha. Finally we must take into effect that many SCV quants are looking at the same data and pickling the same SCV stocks to buy (distorting their prices in the wrong direction and reducing the very thing they seek, value). So after it's all said and done, how is the investor in the average SCV value fund going to do? That is the important question, not whether there is academic research for cost-free long/short portfolios, or convincing narratives for the existence of the SCV premium.

Academics get promotions and often high paying jobs from private factor based investing companies for cranking out convincing factor research. Those who market and sell such products also get handsomely rewarded. There are built-in generous incentives to producing what appears to be a better mouse trap, hence the proliferation of factors and approaches to harvest them. Whether or not the better mouse trap actually works for investors is much less clear than the academic research might suggest.

One final note. All premiums except one are based on cost-free long/short portfolios which, as I have pointed out, is an invalid measuring tool. The exception? Beta. No short is used to double the the expected return of long only beta, except possibly the risk free rate, T Bills which in real terms inflation-adjusted is rarely more than zero. On the other hand cost-free short equity portfolios is often theoretically quite positive in real terms. So when you're comparing expected returns of the the market versus factors it seems reasonable in order to make a valid comparison to almost double the expected return of the market factor in order to level the playing field. Very few real factor funds use long/short portfolios which says something about how shorting works in the real world after costs. Those that do short tend to have inconsistent results. Shorting works if you know which stocks are going to decline in value by a given date in the future and by how much it will decline. The problem is: no one knows that up front on a consistent basis as QSPIX's recent results demonstrate.

Garland Whizzer
Wow. This is the best post I’ve ever seen detailing why factor investing may not work in real life. Thank you. I did not know FF used long/short in their data, that is very important tidbit to know. I concur, costs and the efficient market likely will prevent any known premium to comsistently persist going forward.

User avatar
privatefarmer
Posts: 609
Joined: Mon Sep 08, 2014 2:45 pm

Re: Small Value: 15 years of underperforming

Post by privatefarmer » Sun Nov 25, 2018 2:50 pm

nedsaid wrote:
Sun Nov 25, 2018 12:22 pm
Garland, the better indexes do a good job filtering out the "lottery stocks" the academics rail about. This is why when you use actual indexes to compare factor strategies against indexes that you often don't see the premiums. From all I have been reading, it seems that Quality/Profitability is where we should be looking. Quality/Profitability helps the Size premium as it does the Value premium. If I had to guess, it would probably help the Momentum premium as well. If you set your screening for Small, Value, and Quality, it would seem to me that many "value traps" would be screened out as well.

It might be that it is all about screening out the "anti-factors" more than screening for the factors themselves. Getting those "lottery stocks" and "value traps" out will boost the performance of about anything. I argue that the better indexes work well because the "lottery stocks" are just not there and that many of the "value traps" are not there either.

I think what is happening is that a single factor doesn't work so well anymore. You have to combine at least two now maybe three, or so it seems. The markets are probably getting more efficient even though more and more money is passively invested in indexes. Some smart aleck suggested that academics publishing studies has an adverse effect on the factors, if they would just stop publishing perhaps the factors would reappear.
Well that’s sounds easy enough. Just buy the “good” indexes! And the “good” factors like profitability? How about we then put the “good” indexes and factors into funds available in our 401ks? And of course let the institutions and hedge funds get their slice. Everyone wins!

Come on. Can we not all agree that if it were that easy to get higher returns everyone would do it and the “premium” would instantaneously be evaporated or even negative?

User avatar
Robert T
Posts: 2610
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small Value: 15 years of underperforming

Post by Robert T » Sun Nov 25, 2018 2:58 pm

.
These types of discussion are similar to those 6 years ago, 10 years ago, and 15 years ago, and would not be surprised to see similar discussion in 6 years, 10 years, and 15 years time.

As a practitioner (investing my own resources) I have simply stayed the course - same factor load targets over the past 15+ years. And it has not turned so badly – last time I check was ahead of a similar three fund portfolio, and ACWI/bond portfolio over past 3, 5, 10, and 15 years – better than expected. viewtopic.php?p=3986190#p3986190

I am satisfied with the risk-base explanation for the equity, size, and value premiums to believe in an expectation of higher returns over the long-term. Obviously no guarantees that this will in fact happen for any of them.

As I ended the above linked post.

If you are not convinced about the merits of tilting away from the market portfolio, simply hold the market portfolio. If you do tilt away from the market, then stick with it over the long-term (don’t chop and change every 3-5 years). If I recall correctly, Wes Gray said something along the lines of the following – there is not enough time left in his life to produce enough out of sample data to disprove the existence of the value (and momentum) premiums – hence he is in it for the rest of his life. Shorter-term data 3, 5, 10 yrs etc. does not tell you a whole lot about the long-term existence of factor premiums – yet most of the discussions seem dominated by these shorter-term periods.

Again – no guarantees. Each to their own approach.

Robert
.

User avatar
hdas
Posts: 1337
Joined: Thu Jun 11, 2015 8:24 am

[Deleted]

Post by hdas » Sun Nov 25, 2018 3:41 pm

[Deleted]
Last edited by hdas on Wed Jan 29, 2020 12:08 pm, edited 1 time in total.
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

garlandwhizzer
Posts: 2572
Joined: Fri Aug 06, 2010 3:42 pm

Re: Small Value: 15 years of underperforming

Post by garlandwhizzer » Sun Nov 25, 2018 4:15 pm

Robert T wrote:

Some investors may not believe in the value, momentum, and profitability premiums and want to just hold the market - and thats perfectly okay. Others may believe in these premiums and want some targeted exposure to them. FF research can help to do that in a more structured way than random selection based on fund names.

Obviously no guarantees.
1+

I agree with this completely. Robert T is a very knowledgeable and honest student of the market and a true believer in factor approaches. Personally, I am skeptical that factor approaches will produce superior results going forward for investors, but it is possible that they will. I therefore hedge my bets and hold both in US equity, the biggest bet, 75% in TSM, and 25% in SCV. Those who tilt 100% have more faith in academic research as a guide to future market action than I do. Each to his own. I'm not sure who's right. As Taylor says, there is more than one road to Dublin. If factor funds are well chosen, sticking with either factors of market, or a combo of both, through thick or thin is overwhelmingly likely to produce long term success. Nice things about TSM include simplicity, ultra-low cost, absent trading frictions, tax efficiency, wide diversification, no need to juggle multiple funds, plus the fact that it's very hard on a consistent basis to beat the market.

Garland Whizzer

User avatar
Robert T
Posts: 2610
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small Value: 15 years of underperforming

Post by Robert T » Sun Nov 25, 2018 4:16 pm

hdas wrote:
Sun Nov 25, 2018 2:42 pm
Andrew Ang, who went from writing the best book on factors to manage Blackrock's factor division writes: regarding the behavioral underpinnings of the value factor:
Maybe not enough institutions have sufficiently long horizons to effectively
practice value investing. The value effect documented here, though, is different
from the “deep value” practiced by some investors, including Buffet. That requires
five- to ten-year horizons. The book-to-market value effect described here is a
three- to six-month effect. But perhaps even this horizon is too long for most
“long-horizon” investors.
This horizon (3-6 months) seems relevant for strategy evaluation. H
I think he is referring to differing approaches to value => Buffet deep value - buy and hold companies for long-term vs. a book-to-market based index that is frequently re-balanced (changing the companies held). Some value funds re-balance every three months/quarterly, others annually (re: "The book-to-market value effect described here is a three- to six-month effect"). Andrew Ng also indicates that the "The value premium is robust".

Robert
.

User avatar
Robert T
Posts: 2610
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Small Value: 15 years of underperforming

Post by Robert T » Sun Nov 25, 2018 4:44 pm

garlandwhizzer wrote:
Sun Nov 25, 2018 4:15 pm
I agree with this completely. Robert T is a very knowledgeable and honest student of the market and a true believer in factor approaches. Personally, I am skeptical that factor approaches will produce superior results going forward for investors, but it is possible that they will. I therefore hedge my bets and hold both in US equity, the biggest bet, 75% in TSM, and 25% in SCV. Those who tilt 100% have more faith in academic research as a guide to future market action than I do. Each to his own. I'm not sure who's right. As Taylor says, there is more than one road to Dublin. If factor funds are well chosen, sticking with either factors of market, or a combo of both, through thick or thin is overwhelmingly likely to produce long term success. Nice things about TSM include simplicity, ultra-low cost, absent trading frictions, tax efficiency, wide diversification, no need to juggle multiple funds, plus the fact that it's very hard on a consistent basis to beat the market.

Garland Whizzer
FWIW - I have the highest degree of confidence in beta (market exposure), then value, then size - based on historical evidence, for better or worse. This is reflected in my target equity factor loads of 1.0 / 0.4 / 0.2 for Mkt / Value / Size. i.e. I am not 100% small value tilted. That is just my preference, and can understand why some people are 100% small value tilted, and why some are 100% TSM. As I have managed to stay the source over the past 15+ years, I think I have pitched my tent in about the right spot (for me).

Robert
.

User avatar
nedsaid
Posts: 12983
Joined: Fri Nov 23, 2012 12:33 pm

Re: Small Value: 15 years of underperforming

Post by nedsaid » Sun Nov 25, 2018 7:29 pm

privatefarmer wrote:
Sun Nov 25, 2018 2:50 pm
nedsaid wrote:
Sun Nov 25, 2018 12:22 pm
Garland, the better indexes do a good job filtering out the "lottery stocks" the academics rail about. This is why when you use actual indexes to compare factor strategies against indexes that you often don't see the premiums. From all I have been reading, it seems that Quality/Profitability is where we should be looking. Quality/Profitability helps the Size premium as it does the Value premium. If I had to guess, it would probably help the Momentum premium as well. If you set your screening for Small, Value, and Quality, it would seem to me that many "value traps" would be screened out as well.

It might be that it is all about screening out the "anti-factors" more than screening for the factors themselves. Getting those "lottery stocks" and "value traps" out will boost the performance of about anything. I argue that the better indexes work well because the "lottery stocks" are just not there and that many of the "value traps" are not there either.

I think what is happening is that a single factor doesn't work so well anymore. You have to combine at least two now maybe three, or so it seems. The markets are probably getting more efficient even though more and more money is passively invested in indexes. Some smart aleck suggested that academics publishing studies has an adverse effect on the factors, if they would just stop publishing perhaps the factors would reappear.
Well that’s sounds easy enough. Just buy the “good” indexes! And the “good” factors like profitability? How about we then put the “good” indexes and factors into funds available in our 401ks? And of course let the institutions and hedge funds get their slice. Everyone wins!

Come on. Can we not all agree that if it were that easy to get higher returns everyone would do it and the “premium” would instantaneously be evaporated or even negative?
Well, not that hard. The US Total Stock Market is a "good" index. The S&P indexes like the S&P 500, the S&P Mid-Cap 400, and the S&P Small-Cap 600 Index are "good" indexes. Read good things here about the S&P 600 Small-Cap Value Index. The idea is that there is some screening for quality, one thing being is that the company has to have earnings to be included. Conversely, the Russell 2000 has been criticized as a "bad" index, it has been criticized for being easy for managers to beat.

The better constructed indexes, screen out the "lottery stocks" and many of the "value traps." So if you are invested in a well constructed index, you have won half the battle. Most of the bad stocks the academics rail against are just not there. Plus market cap weighting assures that the indexes will reflect the largest and most successful companies, for example just over 50% of the market cap of the US Total Stock Market Index are represented by just 100 stocks. That gives you a good weighting towards Quality. All you have to do is screen out the "anti-factors".
A fool and his money are good for business.

BigJohn
Posts: 1807
Joined: Wed Apr 02, 2014 11:27 pm

Re: Small Value: 15 years of underperforming

Post by BigJohn » Sun Nov 25, 2018 7:52 pm

vineviz wrote:
Sun Nov 25, 2018 2:37 pm
BigJohn wrote:
Sun Nov 25, 2018 2:13 pm
randomguy wrote:
Sat Nov 24, 2018 5:42 pm
So the question isn't about SV performance against growth. It is about why the Small value mutual funds have not been able to match the fama-French portfolios? I am pretty sure if you compared the intersection of the stocks each held, it would be pretty clear with the average small value mutual fund being larger and less valuely.
Let's assume that the FF portfolio performance is real and persistent. The question for any potential investor then becomes how to move from the academic model to real world implementation. If the funds available (or at least under discussion in this thread) aren't small or valuely enough, then it's not surprising that they don't perform per the FF expectations.
I'm not sure the conclusion you've reached (that existing funds "aren't small or valuely enough") is necessarily the right one. All of the inexpensive (ER < 0.5%) small cap value funds with 10 year track records have statistically significant positive factor loadings on both size and value.

The diligence that investors should exercise comes in making sure these funds aren't ALSO exposed to other factors that the investor doesn't desire, and in making sure that when comparing the small cap value funds to other funds or to a benchmark that appropriate comparisons are made.
I was not try to draw that conclusion, I was just using that very often heard comment on all these SV discussions as the starting point for my comments. Since I’m not a factor investor, I don’t have enough knowledge to comment on the validity of that conclusion one way to the other.

cheezit
Posts: 269
Joined: Sat Jul 14, 2018 7:28 pm

Re: Small Value: 15 years of underperforming

Post by cheezit » Mon Nov 26, 2018 11:27 am

Robert T wrote:
Sun Nov 25, 2018 5:46 am
fennewaldaj wrote:
Sat Nov 24, 2018 10:52 pm
Would it be fair to say small growth as it appears in mutual funds is much better investment than small growth as defined by FF? It seems like none of the funds are really investing in the black hole stocks that make FF SG look so bad.
Yes - index funds that call themselves "small growth" (as well as their historical series) show better annualized returns than the FF Small Growth series. I don't think it is attractive for a mutual fund provider to base a fund exclusively on the FF Small Growth series. What would the sales pitch be?

The 'small growth' funds seems to be more a blend of the FF Small Growth and FF Small Neutral series, and likely explains some of the recent (past 15 years) difference in the returns between FF Small Growth and "Small Growth" index funds. Since 2003 the annually rebalanced FF Small neutral series had a higher return than both the annually rebalanced FF Small Growth and FF Small Value Series. As a result a "growth fund" heavily tilted towards FF Small Neutral (e.g. 20% FF Small Growth:80% FF Small Neutral) since 2003 had higher return than the FF Small Value series as observed with "small growth" index funds vs. "small value" index funds.

That all being said, "small value" has still done as well as, or outperformed, "small growth" in the longer historical index series e.g.

Annualized return since index inception:

7/2001 - 10/2018
CRSP Small Value = 10.3
CRSP Small Growth = 8.4

1/1979 - 10/2018
Russell 2000 Value = 12.8
Russell 2000 Growth = 9.8

7/1995 - 10/2018
S&P SmallCap 600 Value = 11.0
S&P SmallCap 600 Growth = 10.9*

*The way the S&P SmallCap 600 "Growth" series kept up with the "Value" series was to itself have a value tilt (+ momentum). The value load on the "Growth" series was +0.18 for this period.

Obviously no guarantees.

Robert
.
As you are probably aware, this analysis is quite endpoint-sensitive. In another thread I looked at small-value vs small-blend, mid-value vs mid-blend and large-value vs large-blend using the iShares S&P 500/400/600 ETFs from their inception to the present, a roughly 18 year period. Value trailed blend over the period in question for all three sizes, and as I recall the results held up whether you were looking at TWRR for a lump sum or MWRR for a typical investor DCAing paycheck contributions.

This exercise also neatly runs around a couple of limitations of other ways of looking at the data: the S&P indices all have a quality filter of sorts, and since we're comparing the value index to the total index for each size we don't have to worry about the growth index not being anti-value.

To me, the problem with the risk and behavior based narratives for the factor premia is that I can also put together a mental model where eg. the sequence of events goes something like

1992: Fama and French publish their data
2000: in the wake of the tech bubble popping, large numbers of investors looking for somewhere else to put their money put FF's ideas (which by this point have made the jump from academic publication to popular literature and can be put into practice with relatively low-cost funds) into practice and pile into small and value, causing them to skyrocket over the next three years
After 2003: the premium has been arbitraged away in US equities
Some time later: ex-US investors jump on the bandwagon to the extent that the premium gets arbitraged away elsewhere

I also question why FF seem to assume that the effects of all factors are linear. Could a second-order component to the size factor explain "Mel's unloved mid-caps", or is the effect which Mr. Lindauer saw caused by something else (he provides several hypotheses), or is the whole thing a historical accident/seeing patterns in the noise?

User avatar
nisiprius
Advisory Board
Posts: 39799
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Small Value: 15 years of underperforming

Post by nisiprius » Mon Nov 26, 2018 12:17 pm

I still wonder why Fama and French divided size into two bins and value into three.
In June of each year t from 1963 to 1991, all NYSE stocks on CRSP are ranked on size (price times shares). The median NYSE size is then used to split NYSE, Amex, and (after 1972) NASDAQ stocks into two groups, small and big (S and B).... We also break NYSE, Amex, and NASDAQ stocks into three book-to- market equity groups based on the breakpoints for the bottom 30% (low), middle 40% (mediurn) and top 30% (high) of the ranked values of BE/FE for NYSE stocks....

Our decision to sort firms into three groups on BE/ME and only two on ME follows the evidence in Fama and French (1992a) that book-to-market equity has a stronger role in average stock returns than size. The splits are arbitrary, however, and we have not searched over alternatives. The hope is that the tests here and in Fama and French (1992b) are not sensitive to these choices. We see no reason to argue that they are.
This is weird. Are you following? a) They call their own choice "arbitrary." b) They "hope" it doesn't make any difference. c) But they didn't look at any alternatives, so they don't know if it makes any difference or not. And, this is the one I like: d) Fama and French "see no reason to argue"--against Fama and French!
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

User avatar
hdas
Posts: 1337
Joined: Thu Jun 11, 2015 8:24 am

[Deleted]

Post by hdas » Mon Nov 26, 2018 12:40 pm

[Deleted]
Last edited by hdas on Wed Jan 29, 2020 12:08 pm, edited 2 times in total.
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

betablocker
Posts: 504
Joined: Mon Jan 11, 2016 1:26 pm

Re: Small Value: 15 years of underperforming

Post by betablocker » Mon Nov 26, 2018 12:54 pm

nisiprius wrote:
Sun Nov 25, 2018 2:41 pm
betablocker wrote:
Sun Nov 25, 2018 9:35 am
...I believe the FF factor returns are long/short. The funds at dfa and vanguard are long only...
Let's be very clear on this point, because now that long-short funds are getting attention it gets confusing.

1) The Fama-French portfolios, like the ones I cited and St. Lutz cited before me, are long-only.
2) The Fama-French factors are long-short.

Up until a few years ago, most factor advocates were claiming great superiority portfolios with factor tilts obtained from long-only funds providing factor exposure through selection and concentration only.

Even today, even funds with names like "pure value" do not attempt to provide access to the pure value factor itself (i.e. using a long-short portfolio). I think that's a little curious. Is it because nobody wants one, or isn't it feasible?

Along the same lines, if nobody has a Fama-French small-cap value factor fund with a long-short portfolio, why doesn't anyone at least offer a "Fama French small-cap value index fund" that tracks the Fama-French small-cap value portfolio? Again: nobody wants one, or not feasible? And, if not, why not?
What are the FF portfolios? DFA? Those are long only funds set up by a company they helped found but not the research they making freely available on French's website. The point is the differences between the performance of the funds and the factor research is the long/short vs long only difference combined with some trading differences they have implemented.

Random Walker
Posts: 4233
Joined: Fri Feb 23, 2007 8:21 pm

Re: Small Value: 15 years of underperforming

Post by Random Walker » Mon Nov 26, 2018 12:58 pm

nisiprius wrote:
Mon Nov 26, 2018 12:17 pm
I still wonder why Fama and French divided size into two bins and value into three.
I’ve wondered that too. In Larry’s factor book, he looks at the CRSP size data by deciles. 1926-2015 the annualized return rises monotonically with decreasing size.

Dave

User avatar
nisiprius
Advisory Board
Posts: 39799
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Small Value: 15 years of underperforming

Post by nisiprius » Mon Nov 26, 2018 2:09 pm

betablocker wrote:
Mon Nov 26, 2018 12:54 pm
...What are the FF portfolios? DFA?...
No, they are the Fama-French portfolios. That is, they are the portfolios used by Fama and French in their 1992 and 1993 papers that are the foundation of factor investing.

They are the portfolios often used by people when making general claims about the performance of small-cap value, etc.

As far as I know, nobody has ever attempted to found a mutual fund or ETF tracking them. Isn't that interesting?

These: Kenneth R. French Data Library

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

marcopolo
Posts: 2787
Joined: Sat Dec 03, 2016 10:22 am

Re: Small Value: 15 years of underperforming

Post by marcopolo » Mon Nov 26, 2018 3:00 pm

nisiprius wrote:
Mon Nov 26, 2018 2:09 pm
That is, they are the portfolios used by Fama and French in their 1992 and 1993 papers that are the foundation of factor investing.

They are the portfolios often used by people when making general claims about the performance of small-cap value, etc.

As far as I know, nobody has ever attempted to found a mutual fund or ETF tracking them. Isn't that interesting?
I don't have a real strong opinion of factor investing, but that is an interesting observation.

For all the strong proponents of factor investing, can someone hazard a guess as to why that is?
I would expect that group to be clamoring for such funds, so they could at least stop having to explain to skeptics how the funds that claim to do factor investing are not showing the expected results because they are not really capturing the factors properly.
Once in a while you get shown the light, in the strangest of places if you look at it right.

columbia
Posts: 2171
Joined: Tue Aug 27, 2013 5:30 am

Re: Small Value: 15 years of underperforming

Post by columbia » Mon Nov 26, 2018 3:03 pm

Not to be cynical, but you’d think that DFA would be doing it, were it feasible.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Small Value: 15 years of underperforming

Post by vineviz » Mon Nov 26, 2018 4:10 pm

marcopolo wrote:
Mon Nov 26, 2018 3:00 pm
nisiprius wrote:
Mon Nov 26, 2018 2:09 pm
That is, they are the portfolios used by Fama and French in their 1992 and 1993 papers that are the foundation of factor investing.

They are the portfolios often used by people when making general claims about the performance of small-cap value, etc.

As far as I know, nobody has ever attempted to found a mutual fund or ETF tracking them. Isn't that interesting?
I don't have a real strong opinion of factor investing, but that is an interesting observation.

For all the strong proponents of factor investing, can someone hazard a guess as to why that is?
I would expect that group to be clamoring for such funds, so they could at least stop having to explain to skeptics how the funds that claim to do factor investing are not showing the expected results because they are not really capturing the factors properly.
The Fama-French portfolios were never intended to be investable indexes. That means they don't attempt to screen out illiquid or untradeable securities nor do they employ any buffering rules to minimize transaction costs.

That said, there are many low-cost funds which have long track records and high correlations with the Fama-French "Small Value" portfolio. DFA Tax-Managed US Targeted Value (DTMVX), iShares S&P Small-Cap 600 Value ETF (IJS), iShares Russell 2000 Value ETF (IWN), and DFA US Small Cap Value (DFSVX) all have correlations of 0.97 or 0.98 with the Fama-French "Small Value" portfolio since the iShares ETFs were introduced in 2000. I suppose it's a matter of subjective opinion, but I'd say that someone who wanted to capture as much of the Fama-French effect as possible has had plenty of good options available to them.

As far as returns go, $100,000 evenly divided between those four funds (two DFA funds and two iShares ETFs) in August 2000 would have grown to $549,432 as of 10/31/2018. The same amount invested in Vanguard 500 Index fund would have grown to $265,238, or less than half of the small cap value portfolio.

https://www.portfoliovisualizer.com/bac ... tion4_1=25
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

cheezit
Posts: 269
Joined: Sat Jul 14, 2018 7:28 pm

Re: Small Value: 15 years of underperforming

Post by cheezit » Mon Nov 26, 2018 4:42 pm

vineviz wrote:
Mon Nov 26, 2018 4:10 pm
marcopolo wrote:
Mon Nov 26, 2018 3:00 pm
nisiprius wrote:
Mon Nov 26, 2018 2:09 pm
That is, they are the portfolios used by Fama and French in their 1992 and 1993 papers that are the foundation of factor investing.

They are the portfolios often used by people when making general claims about the performance of small-cap value, etc.

As far as I know, nobody has ever attempted to found a mutual fund or ETF tracking them. Isn't that interesting?
I don't have a real strong opinion of factor investing, but that is an interesting observation.

For all the strong proponents of factor investing, can someone hazard a guess as to why that is?
I would expect that group to be clamoring for such funds, so they could at least stop having to explain to skeptics how the funds that claim to do factor investing are not showing the expected results because they are not really capturing the factors properly.
The Fama-French portfolios were never intended to be investable indexes. That means they don't attempt to screen out illiquid or untradeable securities nor do they employ any buffering rules to minimize transaction costs.

That said, there are many low-cost funds which have long track records and high correlations with the Fama-French "Small Value" portfolio. DFA Tax-Managed US Targeted Value (DTMVX), iShares S&P Small-Cap 600 Value ETF (IJS), iShares Russell 2000 Value ETF (IWN), and DFA US Small Cap Value (DFSVX) all have correlations of 0.97 or 0.98 with the Fama-French "Small Value" portfolio since the iShares ETFs were introduced in 2000. I suppose it's a matter of subjective opinion, but I'd say that someone who wanted to capture as much of the Fama-French effect as possible has had plenty of good options available to them.

As far as returns go, $100,000 evenly divided between those four funds (two DFA funds and two iShares ETFs) in August 2000 would have grown to $549,432 as of 10/31/2018. The same amount invested in Vanguard 500 Index fund would have grown to $265,238, or less than half of the small cap value portfolio.

https://www.portfoliovisualizer.com/bac ... tion4_1=25
Inspired by this post, I took a look at a few different scenarios. I'd like to title this chart "sequence of returns can matter in the accumulation stage":

TSM essentially matched the performance of the SV-only portfolio over this period for an investor starting out at the beginning of said period, despite the SV-only portfolio having 4.00% higher TWRR over the period! A 1/n large/mid/small port with no value/growth tilt ended up with $40k more in the bank despite a 1.6% lower TWRR than the SV-only portfolio. Weird!

fennewaldaj
Posts: 823
Joined: Sun Oct 22, 2017 11:30 pm

Re: Small Value: 15 years of underperforming

Post by fennewaldaj » Mon Nov 26, 2018 6:38 pm

cheezit wrote:
Mon Nov 26, 2018 4:42 pm


TSM essentially matched the performance of the SV-only portfolio over this period for an investor starting out at the beginning of said period, despite the SV-only portfolio having 4.00% higher TWRR over the period! A 1/n large/mid/small port with no value/growth tilt ended up with $40k more in the bank despite a 1.6% lower TWRR than the SV-only portfolio. Weird!
right. SInce SV outperformed a ton at the beginning you don't get much effect since that is when you have the least money. If you start in say 1994 it would look better. I guess one should expect over or underperformance of these strategies to be fairly modest since one is adding all the time not all at once.

Northern Flicker
Posts: 5079
Joined: Fri Apr 10, 2015 12:29 am

Re: Small Value: 15 years of underperforming

Post by Northern Flicker » Tue Nov 27, 2018 12:00 am

Even today, even funds with names like "pure value" do not attempt to provide access to the pure value factor itself (i.e. using a long-short portfolio). I think that's a little curious. Is it because nobody wants one, or isn't it feasible?
A fund like that would be an alt portfolio diversifier but not a core holding. If you hedge beta away you are swimming upstream against a strong current. Beta is the most important source of equity returns. I think most of the return of a small-cap value portfolio is market beta. Any size or value premium is, in most markets, extra icing on the cake (1977-2000 excepted, where small-cap value exploded from 1977-1983 and seriously underperformed from 1984-early 2000).
Index fund investor since 1987.

g2morrow
Posts: 75
Joined: Thu Jun 07, 2018 7:23 am

Re: Small Value: 15 years of underperforming

Post by g2morrow » Tue Nov 27, 2018 10:22 am

samsdad wrote:
Thu Nov 22, 2018 7:32 pm
<whispers>
small cap blend!
blend for the win!

OkieIndexer
Posts: 430
Joined: Sun Aug 23, 2009 1:10 pm

Re: Small Value: 15 years of underperforming

Post by OkieIndexer » Mon Dec 10, 2018 7:55 pm

stlutz wrote:
Fri Nov 23, 2018 9:42 pm
The discussion points on this thread are taking on an eerie resemblance to the "Final, Definitive" thread on this subject...

viewtopic.php?t=96441
Oh man, the first post in that thread is hilarious. It pretty much sums up the Bogleheads forum in general. :mrgreen:
"In bull markets, people say 'The more risk I take, the greater my return.' But when people aren't afraid of risk, they'll accept risk without being compensated." -Howard Marks, Oaktree Capital

Topic Author
Elysium
Posts: 1985
Joined: Mon Apr 02, 2007 6:22 pm

Re: Small Value: 15 years of underperforming

Post by Elysium » Tue Dec 11, 2018 6:10 pm

In case it wasn't clear, I have owned SCV for 15+ years, and bought few more shares this week. My first investment into SCV was in 2002 with Vanguard SCV Index fund in Roth, later we got access to DFA SV through retirements accounts from 2005 to 2012, and then DFA Targeted Value fund from 2013 onwards. As of now, we are in DFFVX for SCV for about 10%. I am staying the course, while I do question the efficiency of this strategy, making it clear not abandoning it.

DeadPoets
Posts: 125
Joined: Tue Aug 14, 2018 9:18 am

Re: Small Value: 15 years of underperforming

Post by DeadPoets » Sun Jan 06, 2019 2:03 pm

Kenkat wrote:
Fri Nov 23, 2018 9:12 am
To everything (turn, turn, turn)
There is a season (turn, turn, turn)
And a time to every purpose, under heaven

A time for value, a time for growth
A time for large, a time for small
A time for bonds, a time for cash
A time to ask which car brand is better...

In the first half of the 2000s, Small Cap Growth was considered the ugly step-child of asset classes. This one? Poor risk adjusted returns historically, best to just avoid it altogether. And here we are now and it has outperformed. It seems a common trend that the more hated an asset class becomes, the better future performance will be from that point. Not knowing the timing of all of this means patience is a necessary part of any factor investing.
Ha well done.

Post Reply